Merge http://sucs.org/~rohan/git/gfs2-3.0-nmw

* http://sucs.org/~rohan/git/gfs2-3.0-nmw: (24 commits)
  GFS2: Move readahead of metadata during deallocation into its own function
  GFS2: Remove two unused variables
  GFS2: Misc fixes
  GFS2: rewrite fallocate code to write blocks directly
  GFS2: speed up delete/unlink performance for large files
  GFS2: Fix off-by-one in gfs2_blk2rgrpd
  GFS2: Clean up ->page_mkwrite
  GFS2: Correctly set goal block after allocation
  GFS2: Fix AIL flush issue during fsync
  GFS2: Use cached rgrp in gfs2_rlist_add()
  GFS2: Call do_strip() directly from recursive_scan()
  GFS2: Remove obsolete assert
  GFS2: Cache the most recently used resource group in the inode
  GFS2: Make resource groups "append only" during life of fs
  GFS2: Use rbtree for resource groups and clean up bitmap buffer ref count scheme
  GFS2: Fix lseek after SEEK_DATA, SEEK_HOLE have been added
  GFS2: Clean up gfs2_create
  GFS2: Use ->dirty_inode()
  GFS2: Fix bug trap and journaled data fsync
  GFS2: Fix inode allocation error path
  ...

19 files changed, 666 insertions, 1012 deletions

diff --git a/fs/gfs2/acl.c b/fs/gfs2/acl.c
index 34501b64bc4..65978d7885c 100644
--- a/fs/gfs2/acl.c
+++ b/fs/gfs2/acl.c
@@ -82,7 +82,7 @@ static int gfs2_set_mode(struct inode *inode, umode_t mode)
 		iattr.ia_valid = ATTR_MODE;
 		iattr.ia_mode = mode;
 
-		error = gfs2_setattr_simple(GFS2_I(inode), &iattr);
+		error = gfs2_setattr_simple(inode, &iattr);
 	}
 
 	return error;
@@ -160,6 +160,7 @@ out:
 
 int gfs2_acl_chmod(struct gfs2_inode *ip, struct iattr *attr)
 {
+	struct inode *inode = &ip->i_inode;
 	struct posix_acl *acl;
 	char *data;
 	unsigned int len;
@@ -169,7 +170,7 @@ int gfs2_acl_chmod(struct gfs2_inode *ip, struct iattr *attr)
 	if (IS_ERR(acl))
 		return PTR_ERR(acl);
 	if (!acl)
-		return gfs2_setattr_simple(ip, attr);
+		return gfs2_setattr_simple(inode, attr);
 
 	error = posix_acl_chmod(&acl, GFP_NOFS, attr->ia_mode);
 	if (error)
diff --git a/fs/gfs2/aops.c b/fs/gfs2/aops.c
index f9fbbe96c22..4858e1fed8b 100644
--- a/fs/gfs2/aops.c
+++ b/fs/gfs2/aops.c
@@ -663,7 +663,7 @@ static int gfs2_write_begin(struct file *file, struct address_space *mapping,
 	if (&ip->i_inode == sdp->sd_rindex)
 		rblocks += 2 * RES_STATFS;
 	if (alloc_required)
-		rblocks += gfs2_rg_blocks(al);
+		rblocks += gfs2_rg_blocks(ip);
 
 	error = gfs2_trans_begin(sdp, rblocks,
 				 PAGE_CACHE_SIZE/sdp->sd_sb.sb_bsize);
@@ -787,7 +787,6 @@ static int gfs2_stuffed_write_end(struct inode *inode, struct buffer_head *dibh,
 	u64 to = pos + copied;
 	void *kaddr;
 	unsigned char *buf = dibh->b_data + sizeof(struct gfs2_dinode);
-	struct gfs2_dinode *di = (struct gfs2_dinode *)dibh->b_data;
 
 	BUG_ON((pos + len) > (dibh->b_size - sizeof(struct gfs2_dinode)));
 	kaddr = kmap_atomic(page, KM_USER0);
@@ -804,7 +803,6 @@ static int gfs2_stuffed_write_end(struct inode *inode, struct buffer_head *dibh,
 	if (copied) {
 		if (inode->i_size < to)
 			i_size_write(inode, to);
-		gfs2_dinode_out(ip, di);
 		mark_inode_dirty(inode);
 	}
 
@@ -873,10 +871,6 @@ static int gfs2_write_end(struct file *file, struct address_space *mapping,
 		gfs2_page_add_databufs(ip, page, from, to);
 
 	ret = generic_write_end(file, mapping, pos, len, copied, page, fsdata);
-	if (ret > 0) {
-		gfs2_dinode_out(ip, dibh->b_data);
-		mark_inode_dirty(inode);
-	}
 
 	if (inode == sdp->sd_rindex) {
 		adjust_fs_space(inode);
diff --git a/fs/gfs2/bmap.c b/fs/gfs2/bmap.c
index 7878c473ae6..41d494d7970 100644
--- a/fs/gfs2/bmap.c
+++ b/fs/gfs2/bmap.c
@@ -10,6 +10,7 @@
 #include <linux/spinlock.h>
 #include <linux/completion.h>
 #include <linux/buffer_head.h>
+#include <linux/blkdev.h>
 #include <linux/gfs2_ondisk.h>
 #include <linux/crc32.h>
 
@@ -36,11 +37,6 @@ struct metapath {
 	__u16 mp_list[GFS2_MAX_META_HEIGHT];
 };
 
-typedef int (*block_call_t) (struct gfs2_inode *ip, struct buffer_head *dibh,
-			     struct buffer_head *bh, __be64 *top,
-			     __be64 *bottom, unsigned int height,
-			     void *data);
-
 struct strip_mine {
 	int sm_first;
 	unsigned int sm_height;
@@ -273,6 +269,30 @@ static inline __be64 *metapointer(unsigned int height, const struct metapath *mp
 	return ((__be64 *)(bh->b_data + head_size)) + mp->mp_list[height];
 }
 
+static void gfs2_metapath_ra(struct gfs2_glock *gl,
+			     const struct buffer_head *bh, const __be64 *pos)
+{
+	struct buffer_head *rabh;
+	const __be64 *endp = (const __be64 *)(bh->b_data + bh->b_size);
+	const __be64 *t;
+
+	for (t = pos; t < endp; t++) {
+		if (!*t)
+			continue;
+
+		rabh = gfs2_getbuf(gl, be64_to_cpu(*t), CREATE);
+		if (trylock_buffer(rabh)) {
+			if (!buffer_uptodate(rabh)) {
+				rabh->b_end_io = end_buffer_read_sync;
+				submit_bh(READA | REQ_META, rabh);
+				continue;
+			}
+			unlock_buffer(rabh);
+		}
+		brelse(rabh);
+	}
+}
+
 /**
  * lookup_metapath - Walk the metadata tree to a specific point
  * @ip: The inode
@@ -432,12 +452,14 @@ static int gfs2_bmap_alloc(struct inode *inode, const sector_t lblock,
 {
 	struct gfs2_inode *ip = GFS2_I(inode);
 	struct gfs2_sbd *sdp = GFS2_SB(inode);
+	struct super_block *sb = sdp->sd_vfs;
 	struct buffer_head *dibh = mp->mp_bh[0];
 	u64 bn, dblock = 0;
 	unsigned n, i, blks, alloced = 0, iblks = 0, branch_start = 0;
 	unsigned dblks = 0;
 	unsigned ptrs_per_blk;
 	const unsigned end_of_metadata = height - 1;
+	int ret;
 	int eob = 0;
 	enum alloc_state state;
 	__be64 *ptr;
@@ -540,6 +562,15 @@ static int gfs2_bmap_alloc(struct inode *inode, const sector_t lblock,
 			dblock = bn;
 			while (n-- > 0)
 				*ptr++ = cpu_to_be64(bn++);
+			if (buffer_zeronew(bh_map)) {
+				ret = sb_issue_zeroout(sb, dblock, dblks,
+						       GFP_NOFS);
+				if (ret) {
+					fs_err(sdp,
+					       "Failed to zero data buffers\n");
+					clear_buffer_zeronew(bh_map);
+				}
+			}
 			break;
 		}
 	} while ((state != ALLOC_DATA) || !dblock);
@@ -668,76 +699,6 @@ int gfs2_extent_map(struct inode *inode, u64 lblock, int *new, u64 *dblock, unsi
 }
 
 /**
- * recursive_scan - recursively scan through the end of a file
- * @ip: the inode
- * @dibh: the dinode buffer
- * @mp: the path through the metadata to the point to start
- * @height: the height the recursion is at
- * @block: the indirect block to look at
- * @first: 1 if this is the first block
- * @bc: the call to make for each piece of metadata
- * @data: data opaque to this function to pass to @bc
- *
- * When this is first called @height and @block should be zero and
- * @first should be 1.
- *
- * Returns: errno
- */
-
-static int recursive_scan(struct gfs2_inode *ip, struct buffer_head *dibh,
-			  struct metapath *mp, unsigned int height,
-			  u64 block, int first, block_call_t bc,
-			  void *data)
-{
-	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
-	struct buffer_head *bh = NULL;
-	__be64 *top, *bottom;
-	u64 bn;
-	int error;
-	int mh_size = sizeof(struct gfs2_meta_header);
-
-	if (!height) {
-		error = gfs2_meta_inode_buffer(ip, &bh);
-		if (error)
-			return error;
-		dibh = bh;
-
-		top = (__be64 *)(bh->b_data + sizeof(struct gfs2_dinode)) + mp->mp_list[0];
-		bottom = (__be64 *)(bh->b_data + sizeof(struct gfs2_dinode)) + sdp->sd_diptrs;
-	} else {
-		error = gfs2_meta_indirect_buffer(ip, height, block, 0, &bh);
-		if (error)
-			return error;
-
-		top = (__be64 *)(bh->b_data + mh_size) +
-				  (first ? mp->mp_list[height] : 0);
-
-		bottom = (__be64 *)(bh->b_data + mh_size) + sdp->sd_inptrs;
-	}
-
-	error = bc(ip, dibh, bh, top, bottom, height, data);
-	if (error)
-		goto out;
-
-	if (height < ip->i_height - 1)
-		for (; top < bottom; top++, first = 0) {
-			if (!*top)
-				continue;
-
-			bn = be64_to_cpu(*top);
-
-			error = recursive_scan(ip, dibh, mp, height + 1, bn,
-					       first, bc, data);
-			if (error)
-				break;
-		}
-
-out:
-	brelse(bh);
-	return error;
-}
-
-/**
  * do_strip - Look for a layer a particular layer of the file and strip it off
  * @ip: the inode
  * @dibh: the dinode buffer
@@ -752,9 +713,8 @@ out:
 
 static int do_strip(struct gfs2_inode *ip, struct buffer_head *dibh,
 		    struct buffer_head *bh, __be64 *top, __be64 *bottom,
-		    unsigned int height, void *data)
+		    unsigned int height, struct strip_mine *sm)
 {
-	struct strip_mine *sm = data;
 	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
 	struct gfs2_rgrp_list rlist;
 	u64 bn, bstart;
@@ -783,11 +743,6 @@ static int do_strip(struct gfs2_inode *ip, struct buffer_head *dibh,
 	else if (ip->i_depth)
 		revokes = sdp->sd_inptrs;
 
-	if (ip != GFS2_I(sdp->sd_rindex))
-		error = gfs2_rindex_hold(sdp, &ip->i_alloc->al_ri_gh);
-	else if (!sdp->sd_rgrps)
-		error = gfs2_ri_update(ip);
-
 	if (error)
 		return error;
 
@@ -805,7 +760,7 @@ static int do_strip(struct gfs2_inode *ip, struct buffer_head *dibh,
 			blen++;
 		else {
 			if (bstart)
-				gfs2_rlist_add(sdp, &rlist, bstart);
+				gfs2_rlist_add(ip, &rlist, bstart);
 
 			bstart = bn;
 			blen = 1;
@@ -813,7 +768,7 @@ static int do_strip(struct gfs2_inode *ip, struct buffer_head *dibh,
 	}
 
 	if (bstart)
-		gfs2_rlist_add(sdp, &rlist, bstart);
+		gfs2_rlist_add(ip, &rlist, bstart);
 	else
 		goto out; /* Nothing to do */
 
@@ -887,12 +842,82 @@ out_rg_gunlock:
 out_rlist:
 	gfs2_rlist_free(&rlist);
 out:
-	if (ip != GFS2_I(sdp->sd_rindex))
-		gfs2_glock_dq_uninit(&ip->i_alloc->al_ri_gh);
 	return error;
 }
 
 /**
+ * recursive_scan - recursively scan through the end of a file
+ * @ip: the inode
+ * @dibh: the dinode buffer
+ * @mp: the path through the metadata to the point to start
+ * @height: the height the recursion is at
+ * @block: the indirect block to look at
+ * @first: 1 if this is the first block
+ * @sm: data opaque to this function to pass to @bc
+ *
+ * When this is first called @height and @block should be zero and
+ * @first should be 1.
+ *
+ * Returns: errno
+ */
+
+static int recursive_scan(struct gfs2_inode *ip, struct buffer_head *dibh,
+			  struct metapath *mp, unsigned int height,
+			  u64 block, int first, struct strip_mine *sm)
+{
+	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
+	struct buffer_head *bh = NULL;
+	__be64 *top, *bottom;
+	u64 bn;
+	int error;
+	int mh_size = sizeof(struct gfs2_meta_header);
+
+	if (!height) {
+		error = gfs2_meta_inode_buffer(ip, &bh);
+		if (error)
+			return error;
+		dibh = bh;
+
+		top = (__be64 *)(bh->b_data + sizeof(struct gfs2_dinode)) + mp->mp_list[0];
+		bottom = (__be64 *)(bh->b_data + sizeof(struct gfs2_dinode)) + sdp->sd_diptrs;
+	} else {
+		error = gfs2_meta_indirect_buffer(ip, height, block, 0, &bh);
+		if (error)
+			return error;
+
+		top = (__be64 *)(bh->b_data + mh_size) +
+				  (first ? mp->mp_list[height] : 0);
+
+		bottom = (__be64 *)(bh->b_data + mh_size) + sdp->sd_inptrs;
+	}
+
+	error = do_strip(ip, dibh, bh, top, bottom, height, sm);
+	if (error)
+		goto out;
+
+	if (height < ip->i_height - 1) {
+
+		gfs2_metapath_ra(ip->i_gl, bh, top);
+
+		for (; top < bottom; top++, first = 0) {
+			if (!*top)
+				continue;
+
+			bn = be64_to_cpu(*top);
+
+			error = recursive_scan(ip, dibh, mp, height + 1, bn,
+					       first, sm);
+			if (error)
+				break;
+		}
+	}
+out:
+	brelse(bh);
+	return error;
+}
+
+
+/**
  * gfs2_block_truncate_page - Deal with zeroing out data for truncate
  *
  * This is partly borrowed from ext3.
@@ -1031,7 +1056,7 @@ static int trunc_dealloc(struct gfs2_inode *ip, u64 size)
 		sm.sm_first = !!size;
 		sm.sm_height = height;
 
-		error = recursive_scan(ip, NULL, &mp, 0, 0, 1, do_strip, &sm);
+		error = recursive_scan(ip, NULL, &mp, 0, 0, 1, &sm);
 		if (error)
 			break;
 	}
diff --git a/fs/gfs2/dir.c b/fs/gfs2/dir.c
index 1cc2f8ec52a..8ccad2467cb 100644
--- a/fs/gfs2/dir.c
+++ b/fs/gfs2/dir.c
@@ -240,16 +240,15 @@ fail:
 	return error;
 }
 
-static int gfs2_dir_read_stuffed(struct gfs2_inode *ip, char *buf,
-				 u64 offset, unsigned int size)
+static int gfs2_dir_read_stuffed(struct gfs2_inode *ip, __be64 *buf,
+				 unsigned int size)
 {
 	struct buffer_head *dibh;
 	int error;
 
 	error = gfs2_meta_inode_buffer(ip, &dibh);
 	if (!error) {
-		offset += sizeof(struct gfs2_dinode);
-		memcpy(buf, dibh->b_data + offset, size);
+		memcpy(buf, dibh->b_data + sizeof(struct gfs2_dinode), size);
 		brelse(dibh);
 	}
 
@@ -261,13 +260,12 @@ static int gfs2_dir_read_stuffed(struct gfs2_inode *ip, char *buf,
  * gfs2_dir_read_data - Read a data from a directory inode
  * @ip: The GFS2 Inode
  * @buf: The buffer to place result into
- * @offset: File offset to begin jdata_readng from
  * @size: Amount of data to transfer
  *
  * Returns: The amount of data actually copied or the error
  */
-static int gfs2_dir_read_data(struct gfs2_inode *ip, char *buf, u64 offset,
-			      unsigned int size, unsigned ra)
+static int gfs2_dir_read_data(struct gfs2_inode *ip, __be64 *buf,
+			      unsigned int size)
 {
 	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
 	u64 lblock, dblock;
@@ -275,24 +273,14 @@ static int gfs2_dir_read_data(struct gfs2_inode *ip, char *buf, u64 offset,
 	unsigned int o;
 	int copied = 0;
 	int error = 0;
-	u64 disksize = i_size_read(&ip->i_inode);
-
-	if (offset >= disksize)
-		return 0;
-
-	if (offset + size > disksize)
-		size = disksize - offset;
-
-	if (!size)
-		return 0;
 
 	if (gfs2_is_stuffed(ip))
-		return gfs2_dir_read_stuffed(ip, buf, offset, size);
+		return gfs2_dir_read_stuffed(ip, buf, size);
 
 	if (gfs2_assert_warn(sdp, gfs2_is_jdata(ip)))
 		return -EINVAL;
 
-	lblock = offset;
+	lblock = 0;
 	o = do_div(lblock, sdp->sd_jbsize) + sizeof(struct gfs2_meta_header);
 
 	while (copied < size) {
@@ -311,8 +299,6 @@ static int gfs2_dir_read_data(struct gfs2_inode *ip, char *buf, u64 offset,
 			if (error || !dblock)
 				goto fail;
 			BUG_ON(extlen < 1);
-			if (!ra)
-				extlen = 1;
 			bh = gfs2_meta_ra(ip->i_gl, dblock, extlen);
 		} else {
 			error = gfs2_meta_read(ip->i_gl, dblock, DIO_WAIT, &bh);
@@ -328,7 +314,7 @@ static int gfs2_dir_read_data(struct gfs2_inode *ip, char *buf, u64 offset,
 		extlen--;
 		memcpy(buf, bh->b_data + o, amount);
 		brelse(bh);
-		buf += amount;
+		buf += (amount/sizeof(__be64));
 		copied += amount;
 		lblock++;
 		o = sizeof(struct gfs2_meta_header);
@@ -371,7 +357,7 @@ static __be64 *gfs2_dir_get_hash_table(struct gfs2_inode *ip)
 	if (hc == NULL)
 		return ERR_PTR(-ENOMEM);
 
-	ret = gfs2_dir_read_data(ip, (char *)hc, 0, hsize, 1);
+	ret = gfs2_dir_read_data(ip, hc, hsize);
 	if (ret < 0) {
 		kfree(hc);
 		return ERR_PTR(ret);
@@ -1695,7 +1681,6 @@ int gfs2_dir_del(struct gfs2_inode *dip, const struct dentry *dentry)
 	const struct qstr *name = &dentry->d_name;
 	struct gfs2_dirent *dent, *prev = NULL;
 	struct buffer_head *bh;
-	int error;
 
 	/* Returns _either_ the entry (if its first in block) or the
 	   previous entry otherwise */
@@ -1724,22 +1709,15 @@ int gfs2_dir_del(struct gfs2_inode *dip, const struct dentry *dentry)
 	}
 	brelse(bh);
 
-	error = gfs2_meta_inode_buffer(dip, &bh);
-	if (error)
-		return error;
-
 	if (!dip->i_entries)
 		gfs2_consist_inode(dip);
-	gfs2_trans_add_bh(dip->i_gl, bh, 1);
 	dip->i_entries--;
 	dip->i_inode.i_mtime = dip->i_inode.i_ctime = CURRENT_TIME;
 	if (S_ISDIR(dentry->d_inode->i_mode))
 		drop_nlink(&dip->i_inode);
-	gfs2_dinode_out(dip, bh->b_data);
-	brelse(bh);
 	mark_inode_dirty(&dip->i_inode);
 
-	return error;
+	return 0;
 }
 
 /**
@@ -1829,10 +1807,6 @@ static int leaf_dealloc(struct gfs2_inode *dip, u32 index, u32 len,
 	if (error)
 		goto out_put;
 
-	error = gfs2_rindex_hold(sdp, &dip->i_alloc->al_ri_gh);
-	if (error)
-		goto out_qs;
-
 	/*  Count the number of leaves  */
 	bh = leaf_bh;
 
@@ -1847,7 +1821,7 @@ static int leaf_dealloc(struct gfs2_inode *dip, u32 index, u32 len,
 		if (blk != leaf_no)
 			brelse(bh);
 
-		gfs2_rlist_add(sdp, &rlist, blk);
+		gfs2_rlist_add(dip, &rlist, blk);
 		l_blocks++;
 	}
 
@@ -1911,8 +1885,6 @@ out_rg_gunlock:
 	gfs2_glock_dq_m(rlist.rl_rgrps, rlist.rl_ghs);
 out_rlist:
 	gfs2_rlist_free(&rlist);
-	gfs2_glock_dq_uninit(&dip->i_alloc->al_ri_gh);
-out_qs:
 	gfs2_quota_unhold(dip);
 out_put:
 	gfs2_alloc_put(dip);
diff --git a/fs/gfs2/file.c b/fs/gfs2/file.c
index edeb9e80290..5002408dabe 100644
--- a/fs/gfs2/file.c
+++ b/fs/gfs2/file.c
@@ -59,15 +59,24 @@ static loff_t gfs2_llseek(struct file *file, loff_t offset, int origin)
 	struct gfs2_holder i_gh;
 	loff_t error;
 
-	if (origin == 2) {
+	switch (origin) {
+	case SEEK_END: /* These reference inode->i_size */
+	case SEEK_DATA:
+	case SEEK_HOLE:
 		error = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED, LM_FLAG_ANY,
 					   &i_gh);
 		if (!error) {
 			error = generic_file_llseek_unlocked(file, offset, origin);
 			gfs2_glock_dq_uninit(&i_gh);
 		}
-	} else
+		break;
+	case SEEK_CUR:
+	case SEEK_SET:
 		error = generic_file_llseek_unlocked(file, offset, origin);
+		break;
+	default:
+		error = -EINVAL;
+	}
 
 	return error;
 }
@@ -357,8 +366,15 @@ static int gfs2_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf)
 	unsigned int data_blocks, ind_blocks, rblocks;
 	struct gfs2_holder gh;
 	struct gfs2_alloc *al;
+	loff_t size;
 	int ret;
 
+	/* Wait if fs is frozen. This is racy so we check again later on
+	 * and retry if the fs has been frozen after the page lock has
+	 * been acquired
+	 */
+	vfs_check_frozen(inode->i_sb, SB_FREEZE_WRITE);
+
 	gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &gh);
 	ret = gfs2_glock_nq(&gh);
 	if (ret)
@@ -367,8 +383,15 @@ static int gfs2_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf)
 	set_bit(GLF_DIRTY, &ip->i_gl->gl_flags);
 	set_bit(GIF_SW_PAGED, &ip->i_flags);
 
-	if (!gfs2_write_alloc_required(ip, pos, PAGE_CACHE_SIZE))
+	if (!gfs2_write_alloc_required(ip, pos, PAGE_CACHE_SIZE)) {
+		lock_page(page);
+		if (!PageUptodate(page) || page->mapping != inode->i_mapping) {
+			ret = -EAGAIN;
+			unlock_page(page);
+		}
 		goto out_unlock;
+	}
+
 	ret = -ENOMEM;
 	al = gfs2_alloc_get(ip);
 	if (al == NULL)
@@ -388,7 +411,7 @@ static int gfs2_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf)
 		rblocks += data_blocks ? data_blocks : 1;
 	if (ind_blocks || data_blocks) {
 		rblocks += RES_STATFS + RES_QUOTA;
-		rblocks += gfs2_rg_blocks(al);
+		rblocks += gfs2_rg_blocks(ip);
 	}
 	ret = gfs2_trans_begin(sdp, rblocks, 0);
 	if (ret)
@@ -396,21 +419,29 @@ static int gfs2_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf)
 
 	lock_page(page);
 	ret = -EINVAL;
-	last_index = ip->i_inode.i_size >> PAGE_CACHE_SHIFT;
-	if (page->index > last_index)
-		goto out_unlock_page;
+	size = i_size_read(inode);
+	last_index = (size - 1) >> PAGE_CACHE_SHIFT;
+	/* Check page index against inode size */
+	if (size == 0 || (page->index > last_index))
+		goto out_trans_end;
+
+	ret = -EAGAIN;
+	/* If truncated, we must retry the operation, we may have raced
+	 * with the glock demotion code.
+	 */
+	if (!PageUptodate(page) || page->mapping != inode->i_mapping)
+		goto out_trans_end;
+
+	/* Unstuff, if required, and allocate backing blocks for page */
 	ret = 0;
-	if (!PageUptodate(page) || page->mapping != ip->i_inode.i_mapping)
-		goto out_unlock_page;
-	if (gfs2_is_stuffed(ip)) {
+	if (gfs2_is_stuffed(ip))
 		ret = gfs2_unstuff_dinode(ip, page);
-		if (ret)
-			goto out_unlock_page;
-	}
-	ret = gfs2_allocate_page_backing(page);
+	if (ret == 0)
+		ret = gfs2_allocate_page_backing(page);
 
-out_unlock_page:
-	unlock_page(page);
+out_trans_end:
+	if (ret)
+		unlock_page(page);
 	gfs2_trans_end(sdp);
 out_trans_fail:
 	gfs2_inplace_release(ip);
@@ -422,11 +453,17 @@ out_unlock:
 	gfs2_glock_dq(&gh);
 out:
 	gfs2_holder_uninit(&gh);
-	if (ret == -ENOMEM)
-		ret = VM_FAULT_OOM;
-	else if (ret)
-		ret = VM_FAULT_SIGBUS;
-	return ret;
+	if (ret == 0) {
+		set_page_dirty(page);
+		/* This check must be post dropping of transaction lock */
+		if (inode->i_sb->s_frozen == SB_UNFROZEN) {
+			wait_on_page_writeback(page);
+		} else {
+			ret = -EAGAIN;
+			unlock_page(page);
+		}
+	}
+	return block_page_mkwrite_return(ret);
 }
 
 static const struct vm_operations_struct gfs2_vm_ops = {
@@ -551,8 +588,16 @@ static int gfs2_close(struct inode *inode, struct file *file)
  * @end: the end position in the file to sync
  * @datasync: set if we can ignore timestamp changes
  *
- * The VFS will flush data for us. We only need to worry
- * about metadata here.
+ * We split the data flushing here so that we don't wait for the data
+ * until after we've also sent the metadata to disk. Note that for
+ * data=ordered, we will write & wait for the data at the log flush
+ * stage anyway, so this is unlikely to make much of a difference
+ * except in the data=writeback case.
+ *
+ * If the fdatawrite fails due to any reason except -EIO, we will
+ * continue the remainder of the fsync, although we'll still report
+ * the error at the end. This is to match filemap_write_and_wait_range()
+ * behaviour.
  *
  * Returns: errno
  */
@@ -560,30 +605,34 @@ static int gfs2_close(struct inode *inode, struct file *file)
 static int gfs2_fsync(struct file *file, loff_t start, loff_t end,
 		      int datasync)
 {
-	struct inode *inode = file->f_mapping->host;
+	struct address_space *mapping = file->f_mapping;
+	struct inode *inode = mapping->host;
 	int sync_state = inode->i_state & (I_DIRTY_SYNC|I_DIRTY_DATASYNC);
 	struct gfs2_inode *ip = GFS2_I(inode);
-	int ret;
+	int ret, ret1 = 0;
 
-	ret = filemap_write_and_wait_range(inode->i_mapping, start, end);
-	if (ret)
-		return ret;
-	mutex_lock(&inode->i_mutex);
+	if (mapping->nrpages) {
+		ret1 = filemap_fdatawrite_range(mapping, start, end);
+		if (ret1 == -EIO)
+			return ret1;
+	}
 
 	if (datasync)
 		sync_state &= ~I_DIRTY_SYNC;
 
 	if (sync_state) {
 		ret = sync_inode_metadata(inode, 1);
-		if (ret) {
-			mutex_unlock(&inode->i_mutex);
+		if (ret)
 			return ret;
-		}
-		gfs2_ail_flush(ip->i_gl);
+		if (gfs2_is_jdata(ip))
+			filemap_write_and_wait(mapping);
+		gfs2_ail_flush(ip->i_gl, 1);
 	}
 
-	mutex_unlock(&inode->i_mutex);
-	return 0;
+	if (mapping->nrpages)
+		ret = filemap_fdatawait_range(mapping, start, end);
+
+	return ret ? ret : ret1;
 }
 
 /**
@@ -620,135 +669,18 @@ static ssize_t gfs2_file_aio_write(struct kiocb *iocb, const struct iovec *iov,
 	return generic_file_aio_write(iocb, iov, nr_segs, pos);
 }
 
-static int empty_write_end(struct page *page, unsigned from,
-			   unsigned to, int mode)
-{
-	struct inode *inode = page->mapping->host;
-	struct gfs2_inode *ip = GFS2_I(inode);
-	struct buffer_head *bh;
-	unsigned offset, blksize = 1 << inode->i_blkbits;
-	pgoff_t end_index = i_size_read(inode) >> PAGE_CACHE_SHIFT;
-
-	zero_user(page, from, to-from);
-	mark_page_accessed(page);
-
-	if (page->index < end_index || !(mode & FALLOC_FL_KEEP_SIZE)) {
-		if (!gfs2_is_writeback(ip))
-			gfs2_page_add_databufs(ip, page, from, to);
-
-		block_commit_write(page, from, to);
-		return 0;
-	}
-
-	offset = 0;
-	bh = page_buffers(page);
-	while (offset < to) {
-		if (offset >= from) {
-			set_buffer_uptodate(bh);
-			mark_buffer_dirty(bh);
-			clear_buffer_new(bh);
-			write_dirty_buffer(bh, WRITE);
-		}
-		offset += blksize;
-		bh = bh->b_this_page;
-	}
-
-	offset = 0;
-	bh = page_buffers(page);
-	while (offset < to) {
-		if (offset >= from) {
-			wait_on_buffer(bh);
-			if (!buffer_uptodate(bh))
-				return -EIO;
-		}
-		offset += blksize;
-		bh = bh->b_this_page;
-	}
-	return 0;
-}
-
-static int needs_empty_write(sector_t block, struct inode *inode)
-{
-	int error;
-	struct buffer_head bh_map = { .b_state = 0, .b_blocknr = 0 };
-
-	bh_map.b_size = 1 << inode->i_blkbits;
-	error = gfs2_block_map(inode, block, &bh_map, 0);
-	if (unlikely(error))
-		return error;
-	return !buffer_mapped(&bh_map);
-}
-
-static int write_empty_blocks(struct page *page, unsigned from, unsigned to,
-			      int mode)
-{
-	struct inode *inode = page->mapping->host;
-	unsigned start, end, next, blksize;
-	sector_t block = page->index << (PAGE_CACHE_SHIFT - inode->i_blkbits);
-	int ret;
-
-	blksize = 1 << inode->i_blkbits;
-	next = end = 0;
-	while (next < from) {
-		next += blksize;
-		block++;
-	}
-	start = next;
-	do {
-		next += blksize;
-		ret = needs_empty_write(block, inode);
-		if (unlikely(ret < 0))
-			return ret;
-		if (ret == 0) {
-			if (end) {
-				ret = __block_write_begin(page, start, end - start,
-							  gfs2_block_map);
-				if (unlikely(ret))
-					return ret;
-				ret = empty_write_end(page, start, end, mode);
-				if (unlikely(ret))
-					return ret;
-				end = 0;
-			}
-			start = next;
-		}
-		else
-			end = next;
-		block++;
-	} while (next < to);
-
-	if (end) {
-		ret = __block_write_begin(page, start, end - start, gfs2_block_map);
-		if (unlikely(ret))
-			return ret;
-		ret = empty_write_end(page, start, end, mode);
-		if (unlikely(ret))
-			return ret;
-	}
-
-	return 0;
-}
-
 static int fallocate_chunk(struct inode *inode, loff_t offset, loff_t len,
 			   int mode)
 {
 	struct gfs2_inode *ip = GFS2_I(inode);
 	struct buffer_head *dibh;
 	int error;
-	u64 start = offset >> PAGE_CACHE_SHIFT;
-	unsigned int start_offset = offset & ~PAGE_CACHE_MASK;
-	u64 end = (offset + len - 1) >> PAGE_CACHE_SHIFT;
-	pgoff_t curr;
-	struct page *page;
-	unsigned int end_offset = (offset + len) & ~PAGE_CACHE_MASK;
-	unsigned int from, to;
-
-	if (!end_offset)
-		end_offset = PAGE_CACHE_SIZE;
+	unsigned int nr_blks;
+	sector_t lblock = offset >> inode->i_blkbits;
 
 	error = gfs2_meta_inode_buffer(ip, &dibh);
 	if (unlikely(error))
-		goto out;
+		return error;
 
 	gfs2_trans_add_bh(ip->i_gl, dibh, 1);
 
@@ -758,40 +690,31 @@ static int fallocate_chunk(struct inode *inode, loff_t offset, loff_t len,
 			goto out;
 	}
 
-	curr = start;
-	offset = start << PAGE_CACHE_SHIFT;
-	from = start_offset;
-	to = PAGE_CACHE_SIZE;
-	while (curr <= end) {
-		page = grab_cache_page_write_begin(inode->i_mapping, curr,
-						   AOP_FLAG_NOFS);
-		if (unlikely(!page)) {
-			error = -ENOMEM;
-			goto out;
-		}
+	while (len) {
+		struct buffer_head bh_map = { .b_state = 0, .b_blocknr = 0 };
+		bh_map.b_size = len;
+		set_buffer_zeronew(&bh_map);
 
-		if (curr == end)
-			to = end_offset;
-		error = write_empty_blocks(page, from, to, mode);
-		if (!error && offset + to > inode->i_size &&
-		    !(mode & FALLOC_FL_KEEP_SIZE)) {
-			i_size_write(inode, offset + to);
-		}
-		unlock_page(page);
-		page_cache_release(page);
-		if (error)
+		error = gfs2_block_map(inode, lblock, &bh_map, 1);
+		if (unlikely(error))
 			goto out;
-		curr++;
-		offset += PAGE_CACHE_SIZE;
-		from = 0;
+		len -= bh_map.b_size;
+		nr_blks = bh_map.b_size >> inode->i_blkbits;
+		lblock += nr_blks;
+		if (!buffer_new(&bh_map))
+			continue;
+		if (unlikely(!buffer_zeronew(&bh_map))) {
+			error = -EIO;
+			goto out;
+		}
 	}
+	if (offset + len > inode->i_size && !(mode & FALLOC_FL_KEEP_SIZE))
+		i_size_write(inode, offset + len);
 
-	gfs2_dinode_out(ip, dibh->b_data);
 	mark_inode_dirty(inode);
 
-	brelse(dibh);
-
 out:
+	brelse(dibh);
 	return error;
 }
 
@@ -799,7 +722,7 @@ static void calc_max_reserv(struct gfs2_inode *ip, loff_t max, loff_t *len,
 			    unsigned int *data_blocks, unsigned int *ind_blocks)
 {
 	const struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
-	unsigned int max_blocks = ip->i_alloc->al_rgd->rd_free_clone;
+	unsigned int max_blocks = ip->i_rgd->rd_free_clone;
 	unsigned int tmp, max_data = max_blocks - 3 * (sdp->sd_max_height - 1);
 
 	for (tmp = max_data; tmp > sdp->sd_diptrs;) {
@@ -831,6 +754,7 @@ static long gfs2_fallocate(struct file *file, int mode, loff_t offset,
 	int error;
 	loff_t bsize_mask = ~((loff_t)sdp->sd_sb.sb_bsize - 1);
 	loff_t next = (offset + len - 1) >> sdp->sd_sb.sb_bsize_shift;
+	loff_t max_chunk_size = UINT_MAX & bsize_mask;
 	next = (next + 1) << sdp->sd_sb.sb_bsize_shift;
 
 	/* We only support the FALLOC_FL_KEEP_SIZE mode */
@@ -884,11 +808,12 @@ retry:
 			goto out_qunlock;
 		}
 		max_bytes = bytes;
-		calc_max_reserv(ip, len, &max_bytes, &data_blocks, &ind_blocks);
+		calc_max_reserv(ip, (len > max_chunk_size)? max_chunk_size: len,
+				&max_bytes, &data_blocks, &ind_blocks);
 		al->al_requested = data_blocks + ind_blocks;
 
 		rblocks = RES_DINODE + ind_blocks + RES_STATFS + RES_QUOTA +
-			  RES_RG_HDR + gfs2_rg_blocks(al);
+			  RES_RG_HDR + gfs2_rg_blocks(ip);
 		if (gfs2_is_jdata(ip))
 			rblocks += data_blocks ? data_blocks : 1;
 
diff --git a/fs/gfs2/glops.c b/fs/gfs2/glops.c
index da21ecaafcc..78418b4fa85 100644
--- a/fs/gfs2/glops.c
+++ b/fs/gfs2/glops.c
@@ -28,40 +28,55 @@
 #include "trans.h"
 #include "dir.h"
 
+static void gfs2_ail_error(struct gfs2_glock *gl, const struct buffer_head *bh)
+{
+	fs_err(gl->gl_sbd, "AIL buffer %p: blocknr %llu state 0x%08lx mapping %p page state 0x%lx\n",
+	       bh, (unsigned long long)bh->b_blocknr, bh->b_state,
+	       bh->b_page->mapping, bh->b_page->flags);
+	fs_err(gl->g